'''Training result validator for adaptive classification with KU data.

'''

import argparse

import re

from os import listdir

from os.path import join as pjoin

parser = argparse.ArgumentParser(

    description='Training result validator for adaptive classification with KU data')

parser.add_argument('resultpath', type=str,

                    help='Path to folder containing result_adapt*')

args = parser.parse_args()

path = args.resultpath

schemes = ["result_adapt{}".format(i) for i in range(1, 6)]

p = re.compile(r"epochs_s(\d+)_f(\d+)\.csv")

valid = True

for scheme in schemes:

    for i in range(10):

        rate = (i + 1) * 10

        fs = listdir(pjoin(path, scheme, "r{}".format(rate)))

        mat = [p.match(s) for s in fs]

        grp = [[int(n) for n in m.groups()] for m in mat if m is not None]

        if grp:

            subjs, _ = zip(*grp)

        else:

            subjs = []

        subjs = set(subjs)

        if len(subjs) != 54:

            valid = False

            print("Missing result for {} rate {}".format(scheme, rate))

            curr = len(subjs)

            miss_subjs = set(range(1, 55)) - subjs

            print('Missing subjs: ', miss_subjs)

if valid:

    print('No missing result detected, Looking good!!')